The Actuary, the Astronomer, and the Meteorologist

April 17, 2024

David Scharf, CCA President

Those of you who are regular readers of this blog will not be surprised about my love of astronomy and the recommended soundtrack to this month’s blog post.*  Of course, this is very apropos of what has been dubbed the Great North American Eclipse, which took place on April 8, 2024, and I want to share with you some of my thoughts on the eclipse, including a surprise connection to actuarial science.

While total solar eclipses are not rare (on average once every 18 months), each eclipse is only visible in a small area, much of which may be uninhabited (such as the ocean), which makes this month’s total eclipse all the more special. 

In order to see the total eclipse, I needed to travel to the “path of totality” – and the closer to the eclipse centerline the better to maximize the time of totality. But even more important than the centerline is the weather. There may be a great eclipse in the sky, but cloud cover would prevent me from seeing such greatness. The perfect location to maximize the time of totality and the best likelihood for good weather? Dallas, Texas. And it is to Dallas that I travelled to experience this majestic event. But just a few days away from the eclipse, heavy cloud cover was forecasted for Dallas (and for much of the path of totality). And in a bizarre twist of the odds, the weather forecasts were predicting clear skies in Burlington, Vermont (at the north end of the path) – the very place that initially had the lowest likelihood for clear skies! What was I to do?

While our astronomical models can predict the time and place of the solar eclipse with high accuracy many years in advance (NASA’s website lists future eclipses through the year 3000!), our weather models have difficulty with weather accuracy just a few days away (or even less). It was this weather uncertainty that presented a decision dilemma for me. Do I cut my trip to Dallas short and fly back to New Jersey the day before the eclipse, and then make the long drive up to Burlington, Vermont the next morning? Or stay in Dallas and hope for good weather? Not a simple decision, especially with high second-guessing potential at play. I dug into these weather models (there are a lot more than you may suspect) and saw that there were those that had lighter cloud cover predicted (but not very reassuring). I decided to defy the meteorologists and stay in Dallas; my personal forecast was for clear skies.

On the morning of the eclipse, just as the meteorologists predicted, there were a lot of big clouds in the sky - the sun not even visible. This was not looking good. But then, just minutes before the partial eclipse began, the clouds started to slowly lift; by the time of the total eclipse there were no clouds anywhere near the sun – we had clear skies and a perfect view!

We all know the joke about meteorology being the only profession where you can be wrong most of the time and still keep your job - and meteorologists take a lot of flak for this (despite the chaotic nature of weather systems that greatly affect any accurate ability for predictability). As actuaries we also use predictive modeling; and while our modeling may provide results with accuracy on average in the long term, we cannot say that about specific events. We are more like the meteorologists than the astronomers, with the key advantage that we don’t have to predict what will happen the very next day. Except in this case, in Dallas, I did indeed forecast the next day’s weather, and was (rather fortuitously) spot on!

*Note the perhaps not so obvious lunar allusion in the song title “La femme d’argent”.